In March, during the much-anticipated Apple Watch keynote, Apple carved out a few minutes to announce ResearchKit, a system to enable faster and easier healthcare research. The announcement was received positively, some even saying that the announcement was bigger news than the Apple Watch itself! Within a week, one of the early apps, MyHeart Counts, had enrolled 11,000-patients, an enrollment pace and efficiency unheard of in the healthcare community. Without a doubt, Apple has the opportunity to bring the power of numbers to healthcare.
What It Is
ResearchKit is an SDK (Software Development Kit) containing presentation logic and user interface components for gathering healthcare research data on an iPhone. It can gain informed patient consent, present surveys (whether existing clinically-validated surveys, or novel surveys for a particular study), and also use the sensors on the device to do things like measure vocal tremor, conduct a 6-minute talk test, and measure motor reaction time.
Building a new research app with ResearchKit-powered is a fairly standard mobile app development project. The ResearchKit components certainly accelerate the process, however you will still need an iOS developer, and you will need to follow all the usual software development steps of requirements gathering, design, implementation, stabilization, and then releasing through the App Store. The SDK was recently open-sourced on GitHub. Since most of the SDK relates to the user interface, ResearchKit really only helps with iPhone app development, which some have pointed out may give rise to a sample bias.
What It Is Not
More important than the development of the mobile app, though, it all of the infrastructure behind the app that allows the data to be securely transmitted, stored, aggregated, and analyzed by researchers. In systems of this kind, the scale and complexity of the underlying data service layer is usually considerably larger than the user-facing data collection app. This is especially true in healthcare, given the compliance overhead imposed by regulations like HIPAA.
On this dimension, ResearchKit has no immediate answer. Given Apple’s privacy-centric stance on data collection and aggregation and the sensitivity of the data, they are unlikely to build a cloud service offering for ResearchKit. As such, it will be up to individual researchers to build their own systems, at least until other software vendors move in to fill this need.
Much like the proverbial iceberg where only 10% of the whole is visible, Apple’s ResearchKit is a beautiful, if small, slice of user interface that hides a large and complex underlying platform needed to actually deploy healthcare research apps.